Approaches to Meat Safety
Authors: Andrea Capobianco Dondona (Farm4trade) , Giuseppe Marruchella (University of Teramo) , Ole Arne Alvseike (Animalia) , Gunvor Elise Nagel-Alne (Animalia) , Alberto Tiburzi (Farm4trade) , Axel Dønnum (Fatland) , Ercole Del Negro (Farm4trade)
This study presents the first trial under field conditions of a fully automated, artificial intelligence-based technology for scoring pleurisy in slaughtered pigs (ADAL: Automatic Detection of Abattoir Lesions). The aim was to evaluate the feasibility of using an automated system for scoring pleurisy in slaughtered pigs under field conditions, using an AI-based system that employs deep learning algorithms and image processing techniques to analyse images of pig lungs captured during post-mortem examination. The experimental setup demonstrated the potential of AI-based systems to improve disease detection in pig populations and reduce the risk of errors associated with traditional manual scoring methods. A total of 19,029 images were analysed, and pleurisy was detected in 10.24% of the half-carcasses. The majority of the pigs (89.76%) did not exhibit any signs of pleurisy. Among the pigs that exhibited pleurisy, the most common score was 2 (4.7%), followed by scores of 3 (3.84%) and 1 (1.7%). Norway's low prevalence of pleurisy compared to other countries is attributed to its strict import regulations and the absence of several important respiratory pathogens. The results suggest that AI-based technologies could provide a fast and cheap tool to systematically record lesions in slaughtered pigs, supplying useful data to all stakeholders in the pig industry. The development and application of AI-based technologies could deeply modify the professional life of veterinarians, without affecting their key role in implementing the best disease control strategies. The study represents the first step in developing a fully automatic method for real-time and systematic analysis of animal health, production, and welfare parameters at the slaughterhouse. Future studies are expected to refine and improve the accuracy and efficiency of AI-based technologies in detecting and scoring pig lesions, providing invaluable support to farmers and veterinarians in maintaining the health and productivity of pig herds.
Keywords:
How to Cite: Dondona, A. C. , Marruchella, G. , Alvseike, O. A. , Nagel-Alne, G. E. , Tiburzi, A. , Dønnum, A. & Del Negro, E. (2023) “Fully Automatic Scoring of Pleurisy in Slaughtered Pigs: First Trial Under Field Conditions”, SafePork. 14(1). doi: https://doi.org/10.31274/safepork.16350